Open johnataylor opened 6 years ago
this issue covers https://github.com/microsoft/botbuilder-js/issues/898 so closing the more specific issue.
The initial target here will be a sample.
Most likely factoring is a standalone function that returns emoji characters and their positions.
And then, perhaps additionally
The Tokenizer attempts to split input that includes emojis. This code works for the basic emoji characters (thumbs up, down, smile etc.) but fails on the multi-character emojis that include modifiers for things such as FitzPatrick scale skin tone.
Suggest we simplify the Tokenizer in the product code and move the partial implementation of emoji tokenization into sample. (The Tokenizer should already be an extensibility point - though test coverage for that is weak.)
Justification is that the feature is useful however it is a long way off being a complete implementation. Providing the code, as sample, and the extensibility point makes for a far clearer and more supportable platform. (Extended unicode would make a great sample and interesting blob post.)